Ship detection in a large scene SAR image using image uniformity description factor

摘要

Effective features extraction and high speed processing algorithms are key points for large scene SAR image processing, such as ship detection in the ocean. Feature based ship detection methods can use more information about targets and make results robust in different backgrounds. Although texture is an effective and strong feature in SAR images, its large computation time limits its practice in SAR applications. In this paper, a new texture feature named Image Uniformity Description Factor (IUDF) is proposed to extract uniformity information of SAR images and is used to design a texture based large scene SAR image ship detection algorithm in which accuracy and processing time are taken into consideration as two indexes to test the usefulness of IUDF in practical application. Experiment results show that the algorithm proposed using IUDF can be applied for ship detection in SAR images with a high accuracy and processing speed.